Landscape designers play a crucial role in enhancing the quality of living spaces. Traditional design methods, centered around designer experience, are time-consuming and costly, with limitations in design diversity. This paper presents an automatic generation and beautification model for landscape design based on deep learning to improve efficiency and overcome aesthetic limitations. Experimental results show that NLP technology within the model enhances text processing capabilities and reduces data processing errors, while the residual module CNN improves the quality of auto-generated models. Incorporating point cloud technology enhances the depiction of local details in 3D models. Furthermore, the model supports design optimization and style conversion based on textual descriptions, offering diversified design elements for landscape design automation.

Due to the widespread adoption of social networks, image-text comments have become a prevalent mode of emotional expression compared to traditional text descriptions. However, there are currently two major challenges. The first is the question of how to extract rich representations effectively from both text and images, and the second is the question of how to extract cross-modal shared emotion features. This study proposes a multimodal sentiment analysis method based on a deep feature interaction network (DFINet). It leverages word-to-word graphs and deep attention interaction networks (DAIN) to learn text representations effectively from multiple subspaces. Additionally, it introduces a cross-modal attention interaction network to extract cross-modal shared emotion features efficiently. This approach helps alleviate the difficulties associated with acquiring image-text features and representing cross-modal shared emotion features. Experimental results on the Yelp dataset demonstrate the effectiveness of the DFINet method.

Protecting Oroqen folk songs is not only the only way to reproduce Chinese traditional music culture, but also the only way to rebuild its national spirit and enhance national cultural confidence. In view of the modernity problems of Oroqen folk songs in the current process of inheritance and protection, this paper puts forward the management and optimization methods of music audio-visual archives resources under the background of big data. This paper analyzes and discusses the resource management path of folk music audio-visual archives in Oroqen in the era of big data and designs a set of perfect digital music audio-visual archives resource management platform, which can not only facilitate the collection, storage, management, and utilization of paper files and electronic files in archives, but also optimize the retrieval algorithm of archives. The resource allocation algorithm based on Nash equilibrium solution is used to optimize it. The simulation results show that the proposed method reduces the information resource allocation time and improves the demand satisfaction.

Convolutional neural networks (CNNs) are deep learning methods that are utilized in image processing such as image classification and recognition. It has achieved excellent results in various sectors; however, it still lacks rotation invariant and spatial information. To establish whether two images are rotational versions of one other, one can rotate them exhaustively to see if they compare favorably at some angle. Due to the failure of current algorithms to rotate images and provide spatial information, the study proposes to transform color spaces and use the Gabor filter to address the issue. To gather spatial information, the HSV and CieLab color spaces are used, and Gabor is used to orient images at various orientation. The experiments show that HSV and CieLab color spaces and Gabor convolutional neural network (GCNN) improves image retrieval with an accuracy of 98.72% and 98.67% on the CIFAR-10 dataset.

Cardiovascular disease is one of the deadliest diseases in the world. Accurate analysis and prediction for real-time heart disease are highly significant. To address this challenge, a novel IoT-based automated function monitoring system to promote the e-healthcare system is proposed. The proposed remote healthcare monitoring system uses an IoT framework (RHMIoT) using deep learning and auto encoder-based machine learning algorithms to accurately predict the presence of heart disease. The RHMIoT framework contains two phases: the first phase is to monitor the severity level of the heart disease patient in real-time, and the second phase is used in the medical decision support system to predict the accuracy level of heart disease. To train and test the open-access Framingham dataset, various deep learning and auto encoder-based machine learning techniques are used. The proposed system obtains an accuracy of 0.8714% using the auto encoder-based kernel SVM algorithm compared to other approaches.

Falls are the major cause of injuries and death of elders who live alone at home. Various research works have provided the best solution to the fall detection approach during day vision. However, fall occurs more at the night due to many factors such as low or zero lighting conditions, intake of medication/ drugs, frequent urination due to nocturia disease, and slippery restroom. Based on the required factors, an autonomous monitoring system based on night condition has been proposed through retro-reflective stickers pasted on their upper cloth and infrared cameras installed in the living environment of elders. The developed system uses features such as changes in orientation angle and distance between the retro-reflective stickers to identify the human shape and its characteristics for fall identification. Experimental analysis has also been performed on various events of fall and non-fall activities during the night exclusively in the living environment of the elder, and the system achieves an accuracy of 96.2% and fall detection rate of 92.9%.

Knowledge engineering algorithms such as deep learning models have exhibited tremendous success in solving complex problems. However, the linear nature of the neural network is the primary reason for vulnerability to the perturbed samples. Adversarial attacks pose a severe threat to applying deep models, especially while designing safety-critical applications. This work proposes security attacks against neural architectures. In particular, we introduce a novel method to create adversarial samples. First, we propose a differential evolution population resizing scheme, which enlarges the generation of adversarial samples by allowing adversaries to speed the convergence process. The proposed system is a novel self-adaptive population resizing-based adversarial mechanism. The result shows the success rate for targeted attack LeNet(60.07%), Network_in_Network(97%), Wide_ResNet50(99%), Pure CNN (97%), DenseNet (54.11%),ResNet50(51%) and LeNet(85.13%), Network_in_Network(33.37%), WideResnet(24.40%), Pure_CNN(19.96%),DenseNet (63.67%), ResNet (68.00%) for non targeted attacks respectively.

This work is focusing to develop a portable, low-cost remote diagnostic system for developing countries where the current state of health is not in the advanced stage. People with diseases like epilepsy, Alzheimer’s, an extreme turmeric state, or a disorder that makes it difficult to move have been observed. The authors propose a cost-effective remote neurology assessment health care system. To predict epilepsy form electroencephalogram (EEG) signals in real-time. The authors implemented the machine learning model that has been deployed in the raspberry pi micro-controller. The feature extraction stage was carried out in Matlab. The extracted features from the EEG signals were transferred wirelessly to the model deployed in pi raspberry to clearly predict epilepsy and normality cases. The results of the real-time prediction of the trained and deployed model were provided for the remote diagnosis system. The data visualizations can be done on Android/IOS and Matlab desktop.

Cloud Computing is one of the most popular platforms in recent times and has many services to offer. The resources are deployed on the Cloud and are made available to cloud users over high-speed internet connectivity. Many enterprises think of migrating the data or application hosted from one Cloud to another based on the requirements. Migration from one Cloud to another Cloud requires security as it is vital for any data. This article presents a novel secure framework called ‘InterCloudFramework,’ considering well-established criteria to migrate various services across clouds with minimal supervision and interruption. Security is the primary concern to migrate the data among inter-clouds. The study incorporates the Elliptical-Curve Diffie-Hellman algorithm to encrypt the data and Merkle Hash Trees to check the integrity of the data. In addition to security during migration, the framework reduces the migration time of data or applications.

Cloud computing is a vast area which uses the resources cost-effectively. The performance aspects and security are the main issues in cloud computing. Besides, the selection of optimal features and high false alarm rate to maintain the highest accuracy of the testing are also the foremost challenges focused. To solve these issues and to increase the accuracy, an effective cloud IDS using Grasshopper optimization Algorithm (GOA) and Deep belief network (DBN) is proposed in this paper. GOA is used to choose the ideal features from the set of features. Finally, DBN is developed for classification according to their selected feasible features. The introduced IDS is simulated on the Python platform and the performance of the suggested model of deep learning is assessed based on statistical measures named as Precision, detection accuracy, f-measure and Recall. The NSL_KDD, and UNSW_NB15 are the two datasets used for the simulation, and the results showed that the proposed scheme achieved maximum classification accuracy and detection rate.